Can't get rover to work

[sevenperforce]

Building a super-lightweight rover for my No Moar Boosters challenge, and I can't for the life of me get it to work.

No matter where I control it from, no matter whether I use docking or staging view, no matter what key combination I press, I can't get it to go forward or backward using wheel torque. Won't move. Great at braking, though.

What am I doing wrong?

Edited September 12, 2017 by sevenperforce

[Snark]

Out of curiosity, what happens if you mount a small probe core (or a small docking port) on the front of it, and do "control from here"-- i.e. so it's looking forward instead of up-- and then try?

[sevenperforce]

12 minutes ago, Snark said:

Out of curiosity, what happens if you mount a small probe core (or a small docking port) on the front of it, and do "control from here"-- i.e. so it's looking forward instead of up-- and then try?

I set the reaction wheel as the root part and then rotated the existing probe core 90 degrees; nothing happened whatsoever. I've also tried putting Kerbals in the command seats and controlling from there; also nothing.

Well, I shouldn't say nothing. If I set it to regular staging mode and press W, it starts doing front-flips and rips itself apart.

Edited September 12, 2017 by sevenperforce

[bewing]

A couple problems.

For a rover, the best control point is pointing forward. Second best is pointing straight up. Pointing down pretty much guarantees a rover that won't work.

Additionally, your wheels calculate which way to steer based on their location with respect to the center of mass. Your wheels are on your CoM, so they don't know what to do.

 

This works better:

There is an LY05 steerable airplane wheel at the front that you can't see from this angle.

(The reaction wheel is disabled, by default. You can turn it on if you need to flip the rover or something.)

Edited September 12, 2017 by bewing

[sevenperforce]

52 minutes ago, bewing said:

A couple problems.

For a rover, the best control point is pointing forward. Second best is pointing straight up. Pointing down pretty much guarantees a rover that won't work.

Additionally, your wheels calculate which way to steer based on their location with respect to the center of mass. Your wheels are on your CoM, so they don't know what to do.

 

This works better:

There is an LY05 steerable airplane wheel at the front that you can't see from this angle.

(The reaction wheel is disabled, by default. You can turn it on if you need to flip the rover or something.)

Oh, I bet the CoM thing is what was messing me up.

Why turn the reaction wheel off? I was setting it to hold SAS radial so it would stay level.

[bewing]

7 minutes ago, sevenperforce said:

Oh, I bet the CoM thing is what was messing me up.

Why turn the reaction wheel off? I was setting it to hold SAS radial so it would stay level.

Then you'd better mount your probe core pointing up, and you won't be able to control from the command seats. If your control point is pointing "forward" (the optimal direction) and you tell your rover to hold Radial Out, then your rover will flip up so that its nose is pointing at the sky.

You can alternately set a reaction wheel with a forward-pointing  control point to "SAS only", and then set it to Stability Hold. You kinda get the best of both worlds that way.

But the point is that you don't want pilot input to be interpreted as pitch/yaw instructions to the reaction wheel. If you do, that will flip you in an instant.

 

 

 

Edited September 12, 2017 by bewing

[Snark]

1 hour ago, bewing said:

Additionally, your wheels calculate which way to steer based on their location with respect to the center of mass. Your wheels are on your CoM, so they don't know what to do.

Oh, it's CoM?  Interesting.  I made a unicycle, way back when, and it worked just fine with no problems at all, and I was making no attempts to try to keep the wheel off the CoM.  (In fact, I was trying to get it as close to directly under the CoM as possible, so that the reaction wheel wouldn't have to work overtime:

Admittedly, it wasn't using the wheel to steer, just to provide motive power (all steering was via the reaction wheel). 

(more pictures here)

Also, that was quite a few KSP versions ago, dunno if the functionality may have changed since then.

[sevenperforce]

11 hours ago, bewing said:

You can alternately set a reaction wheel with a forward-pointing  control point to "SAS only", and then set it to Stability Hold. You kinda get the best of both worlds that way.

But the point is that you don't want pilot input to be interpreted as pitch/yaw instructions to the reaction wheel. If you do, that will flip you in an instant.

Setting it to SAS only was what did the trick. Now she's rolling in style!

[mk1980]

did you try using the translation controls (IJKL) instead of the default rotation controls (WASD) to move? maybe that could work (haven't tried).

i like that original design. looks like a funny solution.

i guess i'll use a more conventional rover design with 4 wheels (or maybe 3) for the duna challenge using some lightweight parts (small octagonal girders or something) as the frame. will probably make it small enough to fit into a 2.5m service bay. those things have high crash tolerance, too, so the rover should be well protected when the powered "one way" lander hits the ground.

Edited September 13, 2017 by mk1980

[sevenperforce]

43 minutes ago, mk1980 said:

did you try using the translation controls (IJKL) instead of the default rotation controls (WASD) to move? maybe that could work (haven't tried).

i like that original design. looks like a funny solution.

i guess i'll use a more conventional rover design with 4 wheels (or maybe 3) for the duna challenge using some lightweight parts (small octagonal girders or something) as the frame. will probably make it small enough to fit into a 2.5m service bay. those things have high crash tolerance, too, so the rover should be well protected when the powered "one way" lander hits the ground.

Yeah, IJKL doesn't work.